Theories of knowledge organization

1
Theories of knowledge organization theories of
knowledgeKeynote speech given March 19. 2013,
at the 13th Meeting of the German ISKO
(International Society for Knowledge
Organization), Potsdam, 19th to 20th March
2013Birger Hjørland
2
Outline

• 1. Ontological commitment
• 2. Scientific versus bibliographic
  classifications
• 3. The epistemological basis of classifications
• 4. Approaches to knowledge organization
• 4.1 Automatic versus human classification
• 4.2 User-based and cognitive approaches
• 4.3 Facet classifications
• 4.4 Numeric taxonomic approaches
• 4.5 Bibliometric approaches
• 4.6 Domain-analytic classification
• 5. Conclusion

3

• 1. Ontological commitment

4
1. Ontological commitment

• The core issues in knowledge organization are to
  determine semantic relations between concepts and
  to ascribe subjects to documents.
• For example
• To say that a cat is a mammal
• To say that a given document is about cats
• These issues also involve determining the meaning
  of the words cat, mammal, aboutness (and
  further concepts such as species, concepts,
  generic relation etc.)

5
1. Ontological commitment

• Normally non-experts would just say that we know
  what a cat is and that we know that it is a
  mammal. If challenged we might look it up in an
  authoritative source, either a general
  encyclopedia or an authoritative biological
  handbook or ask some experts.
• In the case of cats and mammals, this might be
  safe, but in many other cases there are no
  consensus among experts. In main-stream
  biological systematics major groups of animals
  (such as fishes and reptiles) are no longer
  regarded as valid systematic units.

6
1. Ontological commitment

• Cladistics is a novel classificatory method and
  philosophy adopted by zoologists in the last few
  decades, which has provided a rather turbulent
  state of zoological classification (see Blake,
  2011).
• In brief the new cladistic approach prefers to
  group animals based exclusively on shared
  ancestry rather than on shared features such as
  in the traditional Linnaean system.

7
1. Ontological commitment

• How to classify animals (and even to say what
  a species is both in an abstract and concrete
  sense as in cats) is not something that is simply
  given but it is something which depends on
  background theories (paradigms), and in
  biological systematics we have today different
  competing paradigms.
• This may be hard to understand and to accept,
  because in the traditional epistemology we start
  seeing and from there construe our theories based
  on the given. Here things are turned up-side
  down Our theories determine what we see
  something as!

8
1. Ontological commitment

• The notion of ontological commitment has come to
  prominence in the second half of the twentieth
  century, mainly through the work of Willard Quine
  1908-2000 .
• On Quines view the right guide to what exists
  is science, so that our best guide to what exists
  is our best current scientific theory what
  exists is what acceptance of that theory commits
  us to. (Craig, 1998)

9
1. Ontological commitment

• Of course things may be classified in many ways.
  Animals may, for example, be classified according
  to weight and size, color, sweetness, and
  usefulness for human beings.
• This is often done, for example, in books for
  children, or by authorities pests etc. There is
  no one correct way of classifying things, and
  classifications should always be considered in
  relation to their purposes, and everything is
  always classified in many ways.
• Where does this leave KO as a scholarly field?

10
1. Ontological commitment

• At present, many, perhaps most, current
  bibliographic classifications for mammals reflect
  quite outdated science. The latest edition of
  DDC, for example, arranges mammals in essentially
  the same way as the second edition of 1885
  (Blake, 2011, p. 469)
• Such outdated classifications may still do
  their job well. The library of the Zoological
  Society of London uses its own scheme, devised in
  the 1960s and largely based on the Bliss
  Bibliographic Classification, to classify the
  monographs it holds. The librarian reports that,
  in most cases, her patrons are able to retrieve
  items and browse the collection effectively
  (Blake, 2011, p. 469-470)

11
1. Ontological commitment

• Can it really be true that such outdated
  classifications may still do their job well?
• Might the reason be that library
  classifications do not serve advanced retrieval
  purposes (e.g. within front-end research)? Or
  that libraries and databases do not support the
  dissemination of new knowledge to the general
  public?
• If we have such low level of ambition concerning
  classification systems, is there then a need for
  KO as a scholarly research discipline? (Can the
  kind of work done in revising DDC be said to be
  scholarly?)

12
1. Ontological commitment

• Even if there is no one true classification,
  it does not mean that any classification is as
  good as anyone else.
• If KO is to be taken seriously as a scholarly
  field, it must be based on knowledge about the
  implications of alternative ways of classifying.
• We should not say we know that X is a kind of
  Y. We should say that according to a given view
  X is considered a kind of Y (but according to
  other views and interests X is a kind of Z).
• The scholarly basis for classification is to
  consider the underlying paradigms and make an
  informed choice.

13

• 2. Scientific / scholarly versus
• bibliographic classifications

14
2. Scientific versus bibliographic
classifications

• Mai (2004, p. 41) argued that scientific
  classification of natural objects, and the
  bibliographic classification of the content of a
  document, are distinct .
• I find this understanding harmful because it
  undermines the important relation between subject
  knowledge and bibliographical classification
  (e.g. between knowledge about zoological taxonomy
  and the design of classification systems on
  animals for bibliographic databases).

15
2. Scientific versus bibliographic
classifications

• The way biologists classify living organisms is,
  for example, reflected in bibliographical
  classification schemes such as the UDC (with some
  delay).
• Blake (2011) writes that cladistics is a novel
  classificatory method and philosophy adopted by
  zoologists in the last few decades, which has
  provided a rather turbulent state of zoological
  classification.
• Both scientific classification and bibliographic
  classification are subject to the same kinds of
  theory-dependence, interpretation and
  difficulties (i.e. the paradigm shift from
  Linnaean to cladistic classification).

16
2. Scientific versus bibliographic
classifications

• The same was shown by Ørom in the field of art
  The way works of art are presented in museums,
  the way histories of art are organized and the
  way bibliographical classifications are organized
  are all dependent on the view of art (theory or
  philosophy of art) which has dominated the people
  making the organization.

17
2. Scientific versus bibliographic
classifications

Figure 1 Social values worldviews scholarly
paradigms (After Ørom, 2003, p. 132)
18

• 3. The epistemological basis of classifications

19
3. The epistemological basis of classifications

• Some classifications are based on logic (e.g.,
  that even numbers are numbers). The philosophical
  school of conceptual analysis is an attempt to
  generalize the use of a priory analysis for
  classification (see Hanna, 1998).
• Some classifications are based on empirical
  studies. A drug is classified as, e.g.
  tranquilizer, based on medical experiments.
• Some classifications are based on human
  conventions (e.g. the borders of a country, who
  is a royal person).

20
3. The epistemological basis of classifications

• Some classifications are based on heritage
  (e.g., who belongs to a certain family). The
  cladistics school in biological systematics which
  today is the dominant school is based on this
  principle.
• Some classifications are based on purpose (e.g.
  tools for cooking).
• Some classifications are based on a mixture of
  criteria (e.g., combined logical, empirical,
  historicist and pragmatic criteria)

21
3. The epistemological basis of classifications

• Logical, empirical, historicist and pragmatic
  methods may each have applications for which they
  are especially relevant but each may also be
  generalized and used more widely because of
  traditions and ideologies.
• How do we determine whether one or another
  classification is best? To evaluate a
  classification is to consider the methods by
  which it has been produced and to evaluate the
  logic, empirical studies, knowledge of human
  conventions, the genealogy, and the goals the
  classification is meant to serve.

22
3. The epistemological basis of classifications

• In other words To evaluate a classification is
  to engage in the research which lies behind the
  classification in order to check its validity and
  relevance.
• All research is influenced by epistemological
  norms or commitments. There is no simply correct
  way of doing research or one correct and
  all-encompassing scientific method and also in
  the theory of knowledge consensus is rare. In my
  view versions of pragmatism/activity theory are
  the best candidate for fruitful philosophy of
  enquiry, but this issue is still open and is in a
  somewhat confused condition today.

23
3. The epistemological basis of classifications

• The classical theories of empiricism and
  rationalism are still very much alive and
  influential in contemporary research (although
  mostly unrecognized). These theories have been
  characterized as a trap, and the point here is
  that if we understand their shortcomings, we may
  avoid the trap and do better research leading to
  better classifications. Empiricism and
  rationalism used to be considered the fundamental
  epistemological positions (and their combination
  was tried by the logical positivists in the
  beginning of the 20th century without success).

24
3. The epistemological basis of classifications

• Because of their shortcomings, we need to include
  some alternatives. I consider four theories the
  basic epistemological theories Empiricism,
  rationalism, historicism and pragmatism/activity
  theory.
• But there are many labels in use today, including
  actor-network theory, cognitivism, critical
  rationalism, critical realism, critical theory,
  dialectical materialism/Marxism, feminist
  epistemology, hermeneutics , paradigm theory,
  phenomenology, postmodernism (late modernism),
  semiotics, social epistemology and social
  constructivism.

25
3. The epistemological basis of classifications

• I do not think that all these epistemologies have
  important different implications for KO. As
  pragmatic philosophers say If a theory is of no
  practical consequence it is indifferent and
  trivial.
• The most important thing in the criticism of
  empiricism, rationalism and positivism is
• Knowledge is a social and historical product made
  to serve certain purposes and interests. It is
  important to reconsider knowledge claims in the
  light of new purposes, conditions and interests.
  

26
3. The epistemological basis of classifications

• Example Textbooks like Harter (1986), Lancaster
  (2003), Large, Tedd Hartley (2001), and
  Svenonius (2000) can be characterized as texts
  that solidify the use of technical and managerial
  language in LIS in the sense that they are
  basically how-to books, constantly referring to
  techniques, standards, principles, methods and
  rules. If one's professional knowledge base has
  such texts at its foundation, no critical
  attitude is developed nor demanded because these
  textbooks do not question at all the role of
  information seeking or of knowledge organization
  systems in culture and society.

27
3. The epistemological basis of classifications

• They do not provide students with a language, an
  understanding, a knowledge that make them capable
  of participating in public discourse debating the
  functionality and legitimacy of these systems
  (Andersen, 2005)
• I believe Jack Andersens quote can be
  interpreted as a critical epistemological view of
  KO. (And, by the way, JA is inspired by activity
  theory).
• However, a critical view cannot be separated from
  knowledge about technical aspects.

28

• 4. Approaches to knowledge organization
• 4.1 Automatic versus human classification
• 4.2 User-based and cognitive approaches
• 4.3 Facet classifications
• 4.4 Numeric taxonomic approaches
• 4.5 Bibliometric approaches
• 4.6 Domain-analytic classification

29

• 4.1 Automatic versus human classification

30
4.1 Automatic versus human classification

• In overviews of KO a fundamental difference
  between computer based versus human based
  classification and indexing is often made. In
  Hjørland (2011) I argue, however, that this
  distinction is theoretically unfruitful.
• Both human indexers and programmers are guided by
  their knowledge (theories) which at the deepest
  is connected to their (implicit) theories of
  knowledge (of which the most important are
  (empiricism, rationalism, positivism) on the one
  hand and (historicism, hermeneutics, pragmatism,
  AT) on the other (to be demonstrated below))

31

• 4.2 User-based and cognitive classifications

32
4.2 User-based and cognitive classifications

• User-based and cognitive views have been
  influential in KO since the 1970s. Hjørland
  (2013b) is a critical analysis of this approach.
• If KO addresses questions such as Should document
  A be classified in class X? Is term A synonymous
  with term B? It is difficult to understand how
  the study of users (qua users) can provide the
  knowledge needed to answer such questions.
• It is therefore difficult to understand why
  user-studies have been so popular an approach in
  KO.

33
4.2 User-based and cognitive classifications

• One hypothesis is that this is caused by the
  influence of empiricist/positivist ideals of
  science
• It seems better science to make empirical studies
  of users than to engage in say literary genre
  studies, theories of art or the philosophy of
  biological taxonomy. (It is also much easier to
  avoid difficult theoretical problems by basing KO
  on the study of user-preferences).
• User-based approaches is thus seen as an ideology
  with a problematic basis with roots in
  empiricism.

34

• 4.3 Facet classifications

35
4.3 Facet classifications

• In Hjørland (2013a) I found that the facet
  analytic approach is based on the epistemology of
  rationalism.
• The strength of this approach is its logical
  principles and the way it provides structures in
  knowledge organization systems (KOS).

36
4.3 Facet classifications

• The main weaknesses are
• its lack of empirical basis and
• its speculative ordering of knowledge without
  basis in the development or influence of theories
  and socio-historical studies.
• It seems to be based on the problematic
  assumption that relations between concepts are a
  priori and not established by the development of
  models, theories and laws.
• This tradition thus demonstrates how rationalism
  as a theory of knowledge has influenced KO.

37

• 4.4 Numeric taxonomic approaches

38
4.4 Numeric taxonomic approaches

• Statistical methods such as cluster analysis,
  factor analysis etc. are used in many different
  sciences and on many different kinds of data
  (e.g. for classification of diseases or
  biological organisms).
• They are also used for classifying documents
  (vector space models, latent semantic indexing,
  etc.) and may therefore also be considered an
  approach to KO.
• (Normally Information Retrieval (IR) and KO are
  considered different fields, but I argue that
  they should be considered different approaches).

39
4.4 Numeric taxonomic approaches

• IR is an extremely wide and complex field, and it
  may seem hasty and problematic to go into this
  field in such an overall way as is attempted
  here. However, these techniques are competing
  with other approaches to KO (and seemingly have
  much more success and authority in academia
  today). I therefore feel that we in KO have to
  take numeric taxonomic/IR-approaches very
  seriously, and if we want to make room for other
  approaches, we have to provide convincing
  argumentation about the limits of mainstream IR.

40
4.4 Numeric taxonomic approaches

• In this presentation I shall limit myself to
  approaches based on statistical measures of
  similarity between documents (including between
  queries and documents).
• Such statistical measures look like objective
  science. However, things may be similar in many
  different ways, and the seemingly objective
  nature of these approaches may be illusionary

41
4.4 Numeric taxonomic approaches

• Even in the field of numerical taxonomy, where
  the use of similarity coefficients has been even
  more widespread than in information retrieval,
  Jackson, Somers and Harvey (1989) were moved to
  conclude that the choice of a similarity
  coefficient is largely subjective and often based
  on tradition or on a posteriori criteria such as
  the interpretability of the results, and went
  on to quote Gordon (1987) Human ingenuity is
  quite capable of providing a post hoc
  justification of dubious classifications.
  (Ellis, Furner-Hines Willett, 1993, p. 144)

42
4.4 Numeric taxonomic approaches

• Presented in these terms, the history of
  research into the use of similarity coefficients
  in text retrieval appears to betray a lack of
  progress (Ellis, Furner-Hines Willett, 1993,
  p. 141).
• My claims are 1) that all numeric taxonomic
  approaches need to have criteria on how to choose
  among alternative algorithms 2) that such
  criteria must be based on subject theories (e.g.
  cladistic research) on what should be considered
  similar.

43
4.4 Numeric taxonomic approaches

• It should also be considered that IR is applied
  to collections of descriptions of objects (e.g.
  animals). We therefore have different levels of
  epistemology
• The way the objects are described in the
  literature influenced by various paradigms.
• The way a given collection is made influences
  which paradigms are dominating
• The way a given similarity measure is chosen
  supports given paradigms at the expense of
  others.

44
4.4 Numeric taxonomic approaches

• Without information from substantial theories,
  IR-approaches are based on problematic empiricist
  ideals and must fail.
• (If they are successful after all, it might be
  that subject knowledge has been used indirectly,
  e.g. by relevance-assessments).

45

• 4.5 Bibliometric classifications

46
4.5 Bibliometric classifications

• Methods based on citation analysis (e.g.
  co-citation analysis and bibliometric coupling)
  represent a unique set of approaches, which can
  be used to construe atlases of sciences, provide
  candidate terms for thesauri and many other kinds
  of knowledge organizing systems (KOS).
• What are the principal strengths and limits of
  such methods?

47
4.5 Bibliometric classifications

• In general, it cannot be expected that methods
  based on citation analysis are able to produce
  intellectual maps such as geographical maps,
  biological taxonomies or periodical systems. A
  geographical structure, for example, places
  different regions in a structure that is
  autonomous in relation to the documents that are
  written about those regions. You cannot produce a
  geographical map of Spain by making, for example,
  bibliometric maps of the literature about Spain
  Yet such autonomous structures as maps of Spain
  are often very useful for information retrieval
  about Spain (Hjørland, 2002, p. 452).

48
4.5 Bibliometric classifications

• In Hjørland (submitted) I make the distinction
  between KOS reflecting intellectual KO versus KOS
  reflecting social KO
• The intellectual aspect of KO is knowledge
  organized in concepts, propositions, models,
  theories, and laws. Such intellectual
  organizations are primarily structured via
  relations of explanatory coherence (Thagard,
  1992, p. 9), which are again primarily related to
  questions concerning truth.

49
4.5 Bibliometric classifications

• The social aspect of KO is knowledge organized
  into academic departments, disciplines,
  cooperative networks, administrative bodies etc.
  Such social organizations are primarily
  structured by the social division of labor in
  societies which are again primarily related to
  questions concerning social relevance, authority
  and power.
• I argue that citation-based approaches are by
  nature social organizations, and by implication
  the principal limits of these approaches are tied
  to the relation between intellectual and social
  KO.

50
4.5 Bibliometric classifications

• Bibliographic methods therefore cannot render
  subject knowledge superfluous (but are themselves
  like numeric taxonomy dependent on subject
  knowledge).
• Although bibliometrics is often associated with
  domain-analysis, I here argue for considering
  these approaches separately.

51

• 4.6 Domain-analytic classification

52
4.6 Domain-analytic classification

• The domain-analytic (DA) view first of all
  recognizes the need for subject knowledge in
  classification and indexing.
• A fine domain-analytic study is Blake (2011) who
  demonstrates solid knowledge about zoological
  taxonomy and the competing approaches in the
  field (cladistics, evolutionary taxonomy and the
  Linnaean system).
• He also carefully discusses the relations between
  scientific theory, quasi-taxonomic groupings, the
  specific demands that information retrieval puts
  on classifications (including the principles of
  literary warrant).

53
4.6 Domain-analytic classification

• Finally Blake describes the classifications used
  by biologists in their writings (monographs) and
  reveals the tendency to use conflicting or
  inconsistent classifications (corresponding to
  Øroms (2003) concept bricolage).
• The domain-analytic view is the only view which
  is fully committed to exploring knowledge
  organization in the light of subject knowledge
  and substantial scholarly theories (and their
  epistemological basis).

54
4.6 Domain-analytic classification

• The criticism raised against other approaches to
  KO is made from the theoretical position of
  domain-analysis. In all cases it was found that
  the subject-knowledge perspective of DA cannot be
  dismissed.
• That is not to say that other approaches (e.g.
  IR-approaches) are not the most efficient
  approaches we have today. I am just claiming that
  their theoretical bases are problematic and that
  such approaches might be improved by considering
  domain-analytic perspectives.

55
4.6 Domain-analytic classification

• Have arguments been put forward against
  domain-analysis?
• Not really. Even leading scholars of other
  approaches seem to avoid direct discussions. The
  resistance towards DA seems mostly to be unvoiced
  or unspoken.
• Although DA is gaining influence, there still
  seems to be a silent resistance. Why?

56
4.6 Domain-analytic classification

• Some possible reasons may be
• the need to design concrete systems rather than
  to develop theoretical principles
• the problematic tendency to use techniques, not
  theory to direct scientific practice (the
  so-called law of the instrument)
• The displeasure to be involved in subject
  knowledge (and the fear of loosing ones own
  disciplinary identity)

57

• 5. Conclusion

58
Conclusion

• The necessity of subject knowledge in KO (as in
  the broader field of information science/ library
  and information science) is certainly not a new
  idea. This kind of knowledge has always been
  assumed in high standard libraries and
  bibliographical databases such as the National
  Library of Medicine and the MEDLINE database.
• (In parallel to teaching qualifications the
  higher the level of teaching, the bigger are the
  demands on subject knowledge, at the university
  level research qualifications are demanded).

59
Conclusion

• In spite of this subject knowledge has been and
  still is extremely neglected in KO. My claim is
  that the neglect of the importance of subject
  knowledge has brought forward a crisis in KO, and
  that no real progress can be observed in the
  field.
• There is real progress in digital technologies,
  and these are used to improve KO and IR, but this
  is mainly progress caused by developments outside
  our field.

60
Conclusion

• Where does this place the theory of knowledge in
  KO? The first thing to say is that you cannot
  classify domains on the basis of theories of
  knowledge (or other metadisciplines, including
  genre studies, the sociology of knowledge etc.).
  Our studies have to be based on concrete domains.
• Epistemology is, however, the best general
  background it is possible to educate people
  within in information science. It is the best
  general preparation we can provide people with in
  order to study any domain.

61
Conclusion

• The general lesson I draw from epistemology is
  that knowledge is created by humans for some
  specific purposes and serve some interests better
  than others. Concepts and semantic relations are
  not a priori or neutral, but should be examined
  in relation to their implications for the users
  they are meant to serve.
• I hope that my discussion of existent approaches
  to KO has been able to make this claim probable.

62

• Thank you for your attention!

63
References

• Ammon, Ulrich (1977). Probleme der
  Soziolinguistik. 2. Aufl. Tübingen Niemeyer.
• Andersen, Jack (2005). Information Criticism
  Where is it? Progressive Librarian, no. 25, pp.
  12-22. http//web.archive.org/web/20110611100746/h
  ttp//www.libr.org/pl/PL25_summer2005.pdf
• Anderson, J.D. Pérez-Carballo, J. (2001a). The
  nature of indexing how humans and machines
  analyze messages and texts for retrieval. Part I
  research, and the nature of human indexing.
  Information Processing Management, 37 (2),
  231-254.
• Anderson, J.D. Pérez-Carballo, J. (2001b). The
  nature of indexing how humans and machines
  analyze messages and texts for retrieval. Part
  II machine indexing, and the allocation of human
  versus machine effort. Information Processing
  Management, 37 (2), 255-277

64
References

• Blake, James (2011). Some Issues in the
  Classification of Zoology. Knowledge
  Organization, 38(6), 463-472.
• Broadfield, A. (1946). The philosophy of
  classification. London Grafton.
• Civallero, Edgardo (2011). UDC Biology Revision
  Project First Stage Class 59 Vertebrates
  http//eprints.rclis.org/16450/1/Civallero20-20U
  DC20Biology20Revision20Project20-202011.pdf
• Cooper, R. (2005). Classifying madness A
  philosophical examination of the diagnostic and
  statistical manual of mental disorders. Berlin
  Springer.
• Craig, E. (1998). Ontology. Routledge
  Encyclopedia of Philosophy, Version 1.0, London
  Routledge.

65
References

• Ellis, David Furner-Hines, Jonathan Willett,
  Peter (1993).Measuring the degree of similarity
  between objects in text retrieval systems.
  Perspectives in Information Management, 3(2),
  128-149
• Feinberg, Melanie (2008). Classification as
  Communication Properties and Design. A
  dissertation submitted in partial fulfillment of
  the requirements for the degree of Doctor of
  Philosophy. Washington University of Washington.
  http//www.ischool.utexas.edu/feinberg/Feinberg2
  0dissertation.pdf
• Feinberg, Melanie (2011). Classification as
  Communication Properties and Design. Ann Arbor,
  MI ProQuest, UMI Dissertation Publishing.
• Fox, Melodie J. (2012). Book review of two books
  by Rick Szostak. Knowledge Organization, 39(4),
  300-303.

66
References

• Gordon, A.D. (1987). A review of hierarchical
  classification. Journal of the Royal Statistical
  Society, Series A (General), 150(2), 119-137.
• Hanna, Robert (1998). Conceptual analysis. In
  Routledge Encyclopedia of Philosophy, Version
  1.0, London Routledge.
• Hansson, Joacim. (2006). Knowledge organization
  from an institutional point of view Implications
  for theoretical practical development.
  Progressive Librarian A Journal for Critical
  Studies Progressive Politics in Librarianship,
  27 3143.
• Hetherington, John (2000). Role of theory and
  experimental design in multivariate analysis and
  mathematical modeling. In H.E.A.Tinsley S. D.
  Brown (eds). Handbook of applied multivariate
  statistics and mathematical modeling (pp. 37-63).
  San Diego Academic press.

67
References

• Hjørland, B. (1998a). Information retrieval, text
  composition, and semantics. Knowledge
  Organization, 25(1/2), 16-31.
• Hjørland, B. (1998b). The classification of
  psychology. Knowledge Organization, 25(4),
  162-201
• Hjørland, B. (2002) The Methodology of
  Constructing Classification Schemes. A Discussion
  of the State-of-the-Art. Advances in Knowledge
  Organization, 8, 450-456.
• Hjørland, B. (2008a). Book review of Ereshefsky
  (2007) The Poverty of the Linnaean Hierarchy A
  Philosophical Study of Biological Taxonomy.
  Knowledge Organization, 35(4), 255-259.
• Hjørland, B. (2008b). Core classification theory
  A reply to Szostak. Journal of Documentation,
  64(3), 333-342.

68
References

• Hjørland, B. (2010). The foundation of the
  concept of relevance. Journal of the American
  Society for Information Science and Technology,
  61(2), 217-237.
• Hjørland, B. (2011). The Importance of Theories
  of Knowledge Indexing and Information retrieval
  as an example. Journal of the American Society
  for Information Science and Technology, 62(1),
  72-77.
• Hjørland, B. (2013a). Facet analysis The logical
  approach to knowledge organization. Information
  Processing Management, 545-557.
• Hjørland, B. (2013b). User-based and cognitive
  approaches to knowledge organization A
  theoretical analysis of the research literature.
  Knowledge Organization, 40(1), 11-27.
• Hjørland, B (submitted). Bibliometrics a dynamic
  approach to knowledge organization

69
References

• ISO 2788 (1986). Documentation Guidelines for
  the establishment and development of monolingual
  thesauri. International Organization for
  Standardization.
• Jackson, D.A., Somers, K.M., Harvey, H.H.,
  (1989). Similarity Coefficients Measures of
  Co-Occurrence and Association or Simply Measures
  of Occurrence? The American Naturalist, 133(3),
  436-453.
• Kaplan, Abraham (1964). The conduct of inquiry
  Methodology for Behavioral Science. New York
  Chandler Publishing.
• Mai, Jens-Erik. 2004. Classification in context
  relativity, reality, and representation.
  Knowledge organization 31(1), 3948.

70
References

• Mai, J.-E. (2012). Den Gode Klassifikation (The
  Good Classification). Royal School of Library and
  Information Science, Copenhagen, Denmark, Sept.
  13, 2012. Inaugural lecture for the
  professorship in information studies .
  http//www.youtube.com/watch?vnXLpK0JqRyM
• Mammen, J. (2008). What is a concept? Journal of
  Anthropological Psychology, 19, 2527.
• Moser, Paul K. (1998). A priori. In Routledge
  Encyclopedia of Philosophy, Version 1.0, London
  Routledge.
• Ørom, A. (2003). Knowledge organization in the
  domain of art studies - history, transition and
  conceptual changes. Knowledge organization,
  30(3/4), 128-143.

71
References

• Rescher, Nicholas (1998). Fallibilism. In
  Routledge Encyclopedia of Philosophy, Version
  1.0, London Routledge
• Small, H. G. (1973). Co-citation in the
  relationship between two documents. Journal of
  the American Society for Information Science,
  24(4), 256269.
• Svenonius, E. (2000). The intellectual foundation
  of information organization. Cambridge, MA MIT
  Press.
• Tennis, Joseph T. (2003). Two Axes of Domains for
  Domain Analysis. Knowledge Organization, 30(3/4),
  191-195.
• Thagard, P. (1992). Conceptual revolutions.
  Princeton Princeton University Press.

72
References

• Toulmin, Stephen (1999). Knowledge as shared
  procedures. In Engeström, Yrjö Miettinen,
  Reijo Punamäki, Raija-Leena (eds.) Perspectives
  on Activity Theory (pp. 70-86). Cambridge, UK
  Cambridge University Press.
• Will, Leonard (2009). Glossary of terms relating
  to thesauri and other forms of structured
  vocabulary for information retrieval. Retrieved
  February 16, 2013 from http//www.willpowerinfo.c
  o.uk/glossary.htm
• Wilson, D. E. Reeder, D. M. eds. (2005).
  Mammal species of the world a taxonomic and
  geographic reference. 3rd ed. Baltimore Johns
  Hopkins University Press.
• Wilson, Patrick (1968). Two kinds of power An
  essay on bibliographic control. Berkeley, CA
  University of California Press.

73
References

• Wilson, Patrick (1983). Second-hand knowledge. An
  inquiry into cognitive authority. Westport, Conn.
  Greenwood Press.